Recorded Video Highlight Generation Using Scene Transition Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Users experience poor viewing experience when reviewing recorded videos due to excessive content of no interest, as smartphones fuse large amounts of image and audio streams without effective editing, leading to lengthy and unengaging content.
Innovation Solution
Perform scene analysis and scene transition analysis on recorded videos to delete invalid segments and edit highlight video segments, then fuse them into a single highlight video.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the electronic device fuses image stream and audio stream continuously obtained within a time period into a video stream, then the video recording function is complete and comprehensive, but the video content becomes excessively long and contains too much uninteresting content
Solution Approach 1:
The patent extracts and removes invalid segments (such as scene transitions, picture zooming, quick camera movements, and severe jerkiness) from the continuous video stream, keeping only the valid and interesting content. This extraction process eliminates unnecessary portions while preserving the core valuable content, thereby improving viewing experience without losing essential video information.
Solution Approach 2:
The patent segments the continuous video stream into multiple valid video segments based on scene analysis and scene transition detection. By dividing the video into distinct meaningful segments and removing invalid portions between them, the system creates a condensed video that maintains comprehensive content while reducing overall length and improving viewer engagement.
2Ease of operation
If manual editing is performed on video clips to remove invalid segments and create highlight videos, then the viewing experience is improved, but the editing time and operational complexity increase significantly
Solution Approach 1:
The patent implements automatic scene analysis and scene transition detection algorithms that enable the video processing system to autonomously identify and segment valid video portions without requiring manual user intervention. The system automatically detects scene changes, identifies invalid segments, and assembles highlight videos, thereby eliminating the time-consuming manual editing process while maintaining high-quality output.
Solution Approach 2:
The patent performs scene analysis and scene transition detection during the video recording process itself, preparing the video data structure in advance with marked scene boundaries and validity indicators. This preliminary processing allows the highlight video to be generated quickly after recording without requiring time-consuming post-processing analysis, thus reducing overall editing time.
3Manufacturing precision
If scene analysis and scene transition analysis are performed on recorded videos to identify valid segments, then the video quality is improved by removing invalid content, but the processing complexity and computational resources increase
Solution Approach 1:
The patent applies scene analysis and scene transition detection selectively to identify key invalid segments rather than analyzing every single frame in exhaustive detail. By focusing computational resources on detecting significant scene changes and transitions rather than processing all video data uniformly, the system achieves high video quality improvement while reducing overall processing complexity and resource consumption.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
This application provides a video processing method and a related apparatus, to perform scene analysis and scene transition analysis on a video recorded by a user, to delete an invalid segment from the recorded video, edit a plurality of highlight video segments from the recorded video, and fuse the plurality of highlight video segments into one highlight video. In this way, viewing experience of the video recorded by the user can be improved.